Lost sensory and motor function in neurologically impaired subjects leads to a devastating quality of life. The state of the art in Brin Machine Interfaces (BMIs), which restore the ability of motor impaired individuals to make movements using volitionally controlled neural signals, has reached an impressive state of development in the past decade, but virtually all rely exclusively on visual feedback. However, these developments overlook the critical importance of somatosensory feedback in motor control. The restoration of somatosensation in BMIs is likely to be the most important development in this field. The overarching goal of the work proposed here is to design a somatosensory prosthesis to restore tactile sensation to sensory impaired subjects. Our approach is novel and significant in a number of respects. First, we will compare the efficac of stimulation in subcortical to that in cortical structures in the rodent's whisker---barel system. We expect that subcortical stimulation will benefit from the combination of simpler local signal encoding and the natural processing provided by upstream circuits, while cortical structures will benefit from the relative ease of access. Second, we will compar the efficacy of stimulus trains optimized to produce activity in the primary somatosensory cortex (S1) that closely mimic that of the corresponding naturally driven activity to nes optimized to maximize information transfer to cortex. We will do so in intact and in de-afferented animals during aroused, passive and active sensing states. This novel optimization approach should allow the animal to interpret the artificial stimulation with little t no training and to generalize readily to novel contexts. While the proposal addresses what is arguably one of the most important limitations of existing BMIs, namely the lack of somatosensory feedback, it is driven by solid neural engineering principles that may have overarching impact on neurostimulation and neuromodulation in clinical applications. The PI offers a potent combination of expertise in engineering and neuroscience, has established track records in BMI research, and is therefore uniquely qualified to carry out the work.